Unlock Gemini CLI: Free AI Power, Installation Guide, and Pro Tips

Installation

npm install -g @google/gemini-cli

After installing, change to your project directory and run:

cd your-project/
gemini

Enabling Gemini‑3‑Pro model

Upgrade to the latest version (e.g., 0.16.0) and open the settings with /settings. Enable Preview Features. Free‑tier users must request access at https://goo.gle/geminicli-waitlist-signup and wait for approval.

Practical tips

Reset daily quota

Run /auth and choose “Login with Google” to switch Gmail accounts, which resets the daily quota.

Switch model

The CLI defaults to auto mode. For complex tasks, use /model to select the pro model. Avoid auto and flash modes for better results.

Custom rules and memory

Define project‑wide rules in a GEMINI.md file placed either in the project root or globally at ~/.gemini/GEMINI.md. Use /memory to add entries to the global rule set and /memory show to review or clean them.

Typical sections in GEMINI.md include:

Development Commands

Operation Constraints

Project Architecture

Authentication and Authorization

Development Process

Important Notes

Add context files

Prepend an @ symbol to a path to add a directory or file to the context, but for large projects dragging files from an IDE (e.g., Cursor) into the CLI is more efficient.

Configure MCP tools

Edit ~/.gemini/settings.json to list desired MCP tools. Example configuration and resulting output are shown in the original article.

Reference

Gemini 3 Pro on Gemini CLI: https://github.com/google-gemini/gemini-cli/blob/main/docs/get-started/gemini-3.md